ABSTRACT Antiretroviral therapy blocks HIV replication, but is not curative, as quiescent integrated HIV DNA is not eradicated by classical drugs. HIV reservoirs of latently infected cells persist indefinitely in infected patients even after long-term periods of treatment and cause rapid viral rebound if the treatment is stopped. Our identification of some patients who were able to durably control viremia after therapy discontinuation (Post- Treatment Controllers, PTCs) introduced the concept of sustained HIV remission and provided proof of concept that such status is achievable at least in some patients. HIV remission in PTCs is characterized by the presence of infected cells, although at low levels. This implies that this situation of long-term viral control requires balanced interactions between HIV-1 and the host. Yet, intriguingly, PTCs do not have the genetic, clinical or immunological characteristics that are commonly found in the rare patients able to naturally control HIV-1 without therapeutic intervention (HIV controllers). HIV remission appears to be favored by the early initiation of antiretroviral treatment and its maintenance for several years before discontinuation. We believe that PTCs benefited from an early treatment that limited the establishment of viral reservoirs and allowed optimal maturation/priming and preservation of critical immune responses. However, only a small fraction of early and durably treated patients are able to control infection after treatment interruption. There is therefore an urgent need to identify the precise mechanisms associated with post-treatment control of infection and define biomarkers that predict the outcome of cART discontinuation. For this, it will be essential to better understand (i) the dynamics of the establishment and maintenance of cellular HIV reservoirs, (ii) the interplay between immune responses and infected cells, and (iii) the impact of early cART on these parameters. We will compare immunological and virological parameters in PTCs and in patients on effective long-term cART, initiated during primary or chronic HIV-infection. We will use empirical analyses and mathematical modeling to (i) define correlates of protection that could identify ART-receiving patients with the highest probability to achieve HIV remission; and (ii) provide a detailed characterization of the effect of treatment initiation on the pattern and the size of the viral reservoirs and immune responses and how this impact the dynamics of viral rebound at the time of treatment initiation. To develop this project we will benefit of access to unique samples from patients in ANRS PRIMO and ANRS VISCONTI Cohorts and to the empirical data generated in a large non-human primate study of 66 SIV-infected macaques initiating cART at different times and maintaining the treatment for 2 years before interruption.